Search

Silo bags have been used by Canadian farmers in the last few years to temporarily store cereal grains, pulses, and oilseeds for up to one year. It is difficult to install temperature and moisture cables inside silo bags or conduct sampling because any cutting of the silo bag damages its hermeticity. Mathematical models could be used to predict spoilage during storage, and the accuracy of mathematical models is influenced by soil temperature. In this study, soil temperature models were developed based on the energy balance on the surface of the ground and heat conduction equations (HCE). The surface of the ground was covered with snow during winter and with vegetation during summer. The developed soil models were coupled with the developed three-dimensional transient heat, mass, and momentum transfer models inside silo bags. The developed models were validated using weather data (solar radiation, temperature, relative humidity, wind speed, and snow covering) and temperatures collected inside bulk canola in silo bags with 9.1% and 10.5% moisture contents. The prediction accuracy associated with the HCE models was compared with that associated with Fourier series, which has been used in the literature. The HCE models developed in this study had higher prediction accuracy than the Fourier series. The maximum absolute difference and average absolute difference between the measured and predicted (by the HCE models) canola temperature at 10 cm height of the canola from the bottom of the silo bag were 3.23°C and <1.79°C ±0.04°C, respectively. [ABSTRACT FROM AUTHOR]

A three-dimensional transient heat, mass, and momentum transfer model was developed to predict temperatures and moisture contents of canola stored inside silo bags under Canadian Prairie conditions. The developed model calculated the condensation and production of water and heat generated by the respiration of microorganisms inside silo bags. This model was coupled with the soil temperature model developed in part I. These developed models were validated using weather data as input and measured temperatures and moisture contents inside silo bags filled with canola at 9.1%, 10.5%, or 14.4% initial moisture content (MC). The developed models had a poor prediction of the temperature and moisture content of the 14.4% MC canola because the canola seeds spoiled and clumped together in less than four months. The developed models could explain more than 90% of the measured temperatures inside the silo bags filled with 10.5% MC canola without underestimation or overestimation. The average absolute difference was less than 1.9°C ±0.1°C and 0.7°C ±0.0°C inside silo bags with 9.1% and 10.5% initial MC, respectively. The developed models could explain more than 94% of the measured moisture contents of canola stored inside the silo bags with ≤10.5% initial MC. The average absolute difference between measured and predicted moisture contents of canola was ≤0.4% ±0.0% inside the silo bags filled with 9.1% and 10.5% MC canola. Simulation results showed that condensation on the canola seeds mostly occurred at the boundary of silo bags, and canola inside hot spots might produce ≥2.5 fold of heat production, which was measured under small-scale laboratory conditions. [ABSTRACT FROM AUTHOR]